Abstract
Intelligent vehicles are widely used in logistics handling, agriculture, medical service, industrial production, and other industries, but they were often not smooth enough in planning the path, and the number of turns was large, resulting in high energy consumption. Aiming at the unsmooth path planning problem of four-wheel intelligent vehicle path planning algorithm, this article proposed an improved genetic and ant colony hybrid algorithm, and the physical model of intelligent vehicle was established. This article first improved ant colony optimization algorithm about heuristic function with the adaptive change of evaporation factor. Then, it improved the genetic algorithm on fitness function, adaptive adjustment of crossover factor, and mutation factor. Last, this article proposed the improved hybrid algorithm with the addition of a deletion operator, adoption of an elite retention strategy, and addition of suboptimal solutions obtained from the improved ant colony algorithm to improved genetic algorithm to obtain optimized new populations. Simulation and physical experiments show that the improved hybrid algorithm is effective. Compared with the traditional hybrid algorithm, the improved hybrid algorithm reduced by 46% in the average number of iterations and 75% in the average number of turns in a simple grid.
Introduction
Path planning is the cornerstone of autonomous navigation systems. While finding the shortest path is a common objective, the "smoothness" of that path is equally vital for physical vehicles. Frequent sharp turns consume more energy, increase mechanical wear, and slow down mission execution. Standard algorithms like Genetic Algorithms (GA) are effective for global searches but often settle for local optima, while Ant Colony Optimization (ACO) can be slow to converge due to its initial "blind" search phase. This research introduces a hybrid approach designed to produce smoother, more efficient paths with faster convergence.
Methodology
The proposed Improved Hybrid Algorithm (IHA) integrates several key enhancements to both ACO and GA frameworks:
Improved Ant Colony Optimization (ACO)
To accelerate initial search speeds, the ACO’s heuristic function was modified to include an "attractive potential" toward the target, providing ants with directional guidance from the start. Furthermore, an adaptive pheromone evaporation rate was implemented—starting high to encourage exploration and decreasing over time to lock in optimal paths.
Improved Genetic Algorithm (GA)
The GA was refined with a multi-objective fitness function that evaluates path length, turning angles, and safety. A critical addition is the deletion operator: if a direct line between two non-adjacent points in a path is obstacle-free, the intermediate point is removed, instantly smoothing the trajectory and reducing unnecessary turns.
Hybrid Integration
The IHA employs a "handshake" strategy: the improved ACO first identifies a set of high-quality suboptimal solutions. These are then injected into the initial population of the improved GA, ensuring the evolutionary process starts with strong "genetic material" rather than random guesses.
Experimental Setup and Results
The algorithm was tested in a MATLAB 2018b environment and validated with a physical four-wheeled vehicle in a 6x6 meter grid. To ensure safe navigation, obstacles were modeled using an expansion treatment that accounts for the vehicle's radius and a reserved safety distance.
Key Results:
- Computational Efficiency: The IHA reduced the average number of iterations by 46–47% in both simple and complex grid environments.
- Path Smoothness: In simple maps, the number of turns was reduced by 75% compared to traditional hybrid algorithms. In complex maps, turns were reduced by 21%.
- Physical Validation: In real-world trials, the IHA produced a path with only 2 turns, whereas standard GA and traditional hybrid algorithms required 6 and 5 turns, respectively.
Conclusion
The proposed improved hybrid algorithm effectively balances path length and smoothness, making it highly suitable for energy-efficient autonomous vehicle navigation. By injecting ACO-derived solutions into a GA framework and utilizing a deletion operator, the system achieves faster convergence and superior path quality. Future research will focus on integrating dynamic obstacle avoidance to handle unpredictable real-world environments.
Summary updated